# Installing libraries
library(TAM) # for Rasch modeling
library(WrightMap) # to build item-person (Wright) maps
library(lsr) # to calculate Cohen's d
library(DescTools) # to calculate eta-squared
library(Rmisc) # to calculate summary statistics
library(tidyverse) # for plotting
library(cowplot) # for image building
library(magick) # for image processing
library(knitr) # for knitting
# Create logo banner for plots
earthlab_orig <- image_read(path = "earth-lab-logo-white.png") %>%
image_scale("x80")
twitter_orig <- image_read(path = "plot-footer-twitter.png") %>%
image_scale("x70")
black_banner <- image_read(path = "black-banner.png")
earthlab_logo <- image_composite(image_scale(black_banner, "1000x100"), earthlab_orig, offset = "+30+10")
twitter_logo <- image_composite(image_scale(black_banner, "1000x100"), twitter_orig, offset = "+540+15")
logo <- image_append(image_scale(c(earthlab_logo, twitter_logo)), stack = FALSE)
logo
Participant Demographics
A total of 53 consenting participants provided demographic information related to gender, race & ethnicity through a series of items included on the pre-program survey instrument, administered prior to the start of the technical workshops.
Raw Responses (%)
Python Skills
Data Science Skills
Data Science Communication
Data Science Practices
Science Identity
Rasch Modeling (Wright Maps)
Python Skills
Data Science Skills
Data Science Communication
Data Science Practices
Science Identity
Analysis of Variance
##
## Call:
## lm(formula = Ability ~ Trial + Dimension + Cohort + Cohort *
## Dimension, data = abil_trial_dimension_all6)
##
## Residuals:
## Min 1Q Median 3Q Max
## -5.228 -1.128 0.032 0.982 5.022
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 2.75115 0.32182 8.549
## TrialBefore -2.93001 0.16283 -17.994
## DimensionData Science Practices -0.05436 0.44032 -0.123
## DimensionData Science Skills 0.16079 0.44032 0.365
## DimensionPython Skills 2.11823 0.44032 4.811
## DimensionScience Identity -0.12235 0.44032 -0.278
## CohortYear 2 -0.09295 0.46503 -0.200
## CohortYear 3 -0.13972 0.42259 -0.331
## DimensionData Science Practices:CohortYear 2 -0.24793 0.65766 -0.377
## DimensionData Science Skills:CohortYear 2 0.32802 0.66517 0.493
## DimensionPython Skills:CohortYear 2 0.72306 0.66517 1.087
## DimensionScience Identity:CohortYear 2 -0.59235 0.66517 -0.891
## DimensionData Science Practices:CohortYear 3 -0.12938 0.59763 -0.216
## DimensionData Science Skills:CohortYear 3 0.12335 0.60141 0.205
## DimensionPython Skills:CohortYear 3 0.70361 0.60141 1.170
## DimensionScience Identity:CohortYear 3 -0.62037 0.60141 -1.032
## Pr(>|t|)
## (Intercept) < 2e-16 ***
## TrialBefore < 2e-16 ***
## DimensionData Science Practices 0.902
## DimensionData Science Skills 0.715
## DimensionPython Skills 2.05e-06 ***
## DimensionScience Identity 0.781
## CohortYear 2 0.842
## CohortYear 3 0.741
## DimensionData Science Practices:CohortYear 2 0.706
## DimensionData Science Skills:CohortYear 2 0.622
## DimensionPython Skills:CohortYear 2 0.278
## DimensionScience Identity:CohortYear 2 0.374
## DimensionData Science Practices:CohortYear 3 0.829
## DimensionData Science Skills:CohortYear 3 0.838
## DimensionPython Skills:CohortYear 3 0.243
## DimensionScience Identity:CohortYear 3 0.303
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.761 on 452 degrees of freedom
## Multiple R-squared: 0.5314, Adjusted R-squared: 0.5158
## F-statistic: 34.17 on 15 and 452 DF, p-value: < 2.2e-16## Analysis of Variance Table
##
## Response: Ability
## Df Sum Sq Mean Sq F value Pr(>F)
## Trial 1 1004.44 1004.44 323.7867 <2e-16 ***
## Dimension 4 563.71 140.93 45.4291 <2e-16 ***
## Cohort 2 1.34 0.67 0.2158 0.8060
## Dimension:Cohort 8 20.59 2.57 0.8295 0.5769
## Residuals 452 1402.18 3.10
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1## eta.sq eta.sq.part
## Trial 0.3356796615 0.4173656492
## Dimension 0.1883110526 0.2866604141
## Cohort 0.0004473933 0.0009538292
## Dimension:Cohort 0.0068797627 0.0144690270t-testing & Cohen’s d
Data Science Communication
##
## Welch Two Sample t-test
##
## data: comfortpost$Ability and comfortpre$Ability
## t = 10.27, df = 87.882, p-value < 2.2e-16
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## 1.950954 2.887157
## sample estimates:
## mean of x mean of y
## 2.415192354 -0.003863267## [1] 2.096379Data Science Practices
##
## Welch Two Sample t-test
##
## data: confpost$Ability and confpre$Ability
## t = 7.5006, df = 92.239, p-value = 3.837e-11
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## 1.531676 2.634901
## sample estimates:
## mean of x mean of y
## 2.074584526 -0.008704105## [1] 1.531062Science Identity
##
## Welch Two Sample t-test
##
## data: idpost$Ability and idpre$Ability
## t = 3.3174, df = 82.505, p-value = 0.001353
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## 0.6008267 2.4004040
## sample estimates:
## mean of x mean of y
## 1.43789668 -0.06271865## [1] 0.6917187Data Science Skills
##
## Welch Two Sample t-test
##
## data: techconfpost$Ability and techconfpre$Ability
## t = 10.45, df = 89.127, p-value < 2.2e-16
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## 2.431845 3.573682
## sample estimates:
## mean of x mean of y
## 3.0032369769 0.0004732148## [1] 2.179052Python Skills
##
## Welch Two Sample t-test
##
## data: techcomfortpost$Ability and techcomfortpre$Ability
## t = 15.426, df = 89.051, p-value < 2.2e-16
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## 4.968754 6.437964
## sample estimates:
## mean of x mean of y
## 6.641079 0.937720## [1] 3.216642
Growth across the five dimensions
Splitting by cohort
